Antiretroviral nucleoside analogs used in highly active antiretroviral therapy (HAART) are associated with a variety of tissue toxicities associated with mitochondrial DNA depletion, suggesting a block in mt-DNA replication. Since the triphosphate-forms of these analogs variably inhibit mt-DNA polymerase, this enzyme has been promoted as the major target of toxicity associated with HAART. However, AZT (Zidovudine), a widely used drug in AIDS therapy, does not fit this hypothesis very well. AZT is a weak inhibitor of the mitochondrial polymerase and it is not readily converted to the triphosphate form. In recent work form our laboratory we have shown that AZT is a potent inhibitor of thymidine phosphorylation in isolated heart, liver, and brain mitochondria, and in the isolated perfused heart. We have recently demonstrated that AZT significantly decreases the TTP pool in the perfused heart. From this work we have proposed an alternative hypothesis in which the pro-drug AZT inhibits thymidine phosphorylation, reducing the TTP pool leading to imbalances in the deoxynucleotide pool that inhibit mitochondrial DNA replication and lead to mitochondrial DNA depletion. We propose to continue to test this hypothesis in several models of AZT toxicity. In the first we propose to use a previously published model of AZT toxicity in 3T3-F224A cells that are being programmed to differentiate into adipocytes. As these models develop AZT toxicity, we will measure the rates of thymidine and AZT phosphorylation, determine the effects of AZT on thymidine phosphorylation, document changes in the deoxynucleotide pools, and in levels of mitochondrial DNA relative to nuclear DNA and measure mRNA levels of select deoxynucleotide salvage and de novo enzymes. Since AZT is a competitive inhibitor of thymidine kinase 2, we will also test the effects of thymidine and uridine on reversing AZT toxicity, and on reversing changes in the deoxynucleotide pools. Additionally, we propose to use a previously published rat model of AZT toxicity in which liver and heart toxicities and mitochondrial DNA depletion have been observed. Here we will measure the same parameters as described above in the cell culture model including the effect of uridine/thymidine supplementation on AZT toxicity. We also propose to test the hypothesis that toxicity of other thymidine analogs, such as d4T is related to the ratio of analog triphosphate to the naturally occurring dNTP, particularly as cells differentiate and become post-mitotic. Nucleoside analogs represent an important component of HAART. The aims of this grant are to obtain a better understanding of the toxic effects of the nucleoside analog drugs used world-wide in the treatment of AIDS. This information is very important in developing strategies for treating these toxicities and in the rational design of new drugs with less toxic side-effects.